satellite constellation cost modeling: an aggregate model · 6/14/2016 · 1 satellite...

1

Satellite Constellation Cost Modeling: An Aggregate Model

Veronica Foreman1, Jacqueline Le Moigne2, Olivier de Weck1

ESTF2016 – Session B1June 14, 2016

1. Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, MA 021392. NASA Goddard Space Flight Center, 8800 Greenbelt Rd, Greenbelt, MD 20771

2

Presentation Outline

•  Introduction –  TAT-C–  VCR Module

•  Historical Context •  Aggregate model formulation

–  Motivation and Justification–  State of the Art –  Limitations of Existing Models

•  Cost Module, Version 1–  Implementation–  Future Revisions –  Impact

•  Conclusions/Future Work

3

Tradespace Analysis Tool for Design of Distributed Missions PI: Dr. Jacqueline Le Moigne, NASA GSFC

(Le Moigne: TAT-C) NNH14ZDA001N-AIST July 11, 2014

COMPETITION SENSITIVE – FOR REVIEW PURPOSES ONLY

1

Scientific/Technical/Management

General Definition of Distributed Spacecraft Missions: A Distributed Spacecraft Mission (DSM) is a mission

that involves multiple spacecraft to achieve one or more common goals.

1. Project Summary Under a changing technological and economic environment, there is growing interest in implementing future NASA Earth Science missions as Distributed Spacecraft Missions (DSM). The objective of our project is to provide a framework that will facilitate the DSM Pre-Phase A investigations, and that will enable to optimize the design of DSM with respect to a-priori Science goals by conducting trades such as “utilizing a homogeneous constellation in which all spacecraft and instrument payload are identical vs. utilizing a heterogeneous constellation by splitting bands over a formation flying vs. utilizing a disaggregated approach or fractionated spacecraft where functional capabilities are distributed among different spacecraft”.

Figure 1 – Our Trade-space Analysis Tool for Constellations, TAT-C, will enable to perform Pre-Phase A trades such as number of satellites, number of planes, orbit altitude, etc. , to optimize the design of a mission aimed at responding to specific Science goals. With a graphically-rich user interface, it could be utilized by

Science Program Managers to optimize their Science mission portfolio or by Scientists designing their missions as a companion tool to Observing System Simulation Experiments (OSSEs)

Other questions investigated with our Trade-space Analysis Tool for Constellations (TAT-C) could be “Which type of constellations should be chosen? How many spacecraft should be included in the constellation? Should they all be in the same plane? If not, how many planes should be utilized? Which design has the best cost/risk value?” We envision the output of our

NRC Earth Science Decadal

Survey (DS)

TAT-C Tradespace

Analysis Tool for Constellations

Pre-Phase A

Design

Science Program Managers

Earth Scientists

Recommended

Missions

Proposed Missions Scien

ce Goals

Recommended

Missions

Optimize Science Mission Portfolio based on Science Goals and DS as a function of Cost and Risk

Observing System

Simulation Experiments

(OSSEs)

New Missions Design

OPTIONS

(1)

(2)

(3)

•  Characteris*cs+•  Cost,+Risk,+etc.+



Average Revisit Time depends on Number of Planes and Number of Satellites per Plane

Average Revisit Time depends on Orbit Altitude and Instrument Elevation

Objectives: •  Provide a framework to perform pre-

Phase A mission analysis of Distributed Spacecraft Missions (DSM)–  Handle multiple spacecraft sharing mission

objectives–  Include sets of smallsats up through

flagships–  Explore tradespace of variables for pre-

defined science, cost and risk goals, and metrics

–  Optimize cost and performance across multiple instruments and platforms vs. one at a time

•  Create an open access toolset which handles specific science objectives and architectures–  Increase the variability of orbit

characteristics, constellation configurations, and architecture types

–  Remove STK licensing restrictions

4

Value, Cost, and Risk Module

•  Addresses the TAT-C Objectives that require cost and risk evaluations; given a satellite constellation architecture, the VCR module will provide estimates of: –  Value, expressed in dollars or utility –  Cost, life cycle cost (RDT&E, manufacturing, launch, operations) –  Risk, profile of the system technical and cost risk

•  VCR Module will enable trades between performance and value/cost/risk more readily

This presentation addresses the need for an automated, integrated cost model for constellation mission design and

the associated cost estimating challenges.

5

Historical Context

Satellite Constellations

Cost Estimation

6

Building an Aggregate Cost Model

•  Motivation –  Constellation architectures require that traditional cost estimating

assumptions be challenged –  Previous work has highlighted the limitations of existing models with

respect to constellation missions

•  Objective–  Develop an automated cost estimating approach for constellation

missions that will help enable early design phase trades–  Build the approach in such a way that it is easily manipulated and

highly transparent

•  Challenges–  Automated cost estimation often results in skepticism–  Model must be able to adapt to technological innovation

7

Selected Approach

•  Interoperable, parametric cost estimating tool to interface with TAT-C–  Parametric estimating allows for a top down approach

•  More appropriate in early stages of design; does not require extensive design decisions

•  Cost Estimating Relationships (CERs) can be easily updated –  Want to leverage existing and trusted techniques and apply

them to Distributed Spacecraft Mission (DSM) and constellation architectures

•  Allow for relative trades between cost and capability–  Early stage mission cost estimates are relative, not absolute

•  Plan to supplement the parametric approach with an analogous cost estimate to ensure model fidelity

8

State-of-the-Art and Existing Literature

•  Many widely accepted cost estimating tools exist, including: –  Unmanned Space Vehicle Cost Model (USCM), Version 10–  Small Satellite Cost Model (SSCM), 2014 Release–  NASA Instrument Cost Model (NICM), Version 7–  QuickCost, Version 6.0–  Programmatic Estimating Tool (PET)

•  Popular references:–  NASA Cost Estimating Handbook, Version 4.0–  Space Mission Analysis and Design, 3rd Edition

•  Previous work has highlighted the limitations of these tools for constellation missions: –  Limitations of traditional cost models for high performance small

satellites, motivating the SSCM [Abramson and Bearden, 1993]–  Small satellite learning curve parameters, COTS components,

technological complexity as they pertain to DSMs [Nag et al.,2014]

9

Model Comparison

Unmanned Space Vehicle Cost Model (USCM, Version 7) and Small Satellite Cost Model (SSCM, 1998) results for TAT-C generated spacecraft

–  Spacecraft are identical, with IR Sensor payloads, except for total mass –  Payload cost differs substantially between the two models

•  Motivation for alternate payload costing approach

10

Implementation

•  VCR Module Cost Routine combines existing models and applies them to DSMs1.  Assesses mission characteristics (e.g. number of spacecraft)2.  Costs spacecraft and payloads appropriately

•  USCM for spacecraft >= 1000kg•  SSCM for spacecraft < 1000kg

–  Future work: Implementing a cube satellite specific model for < 20kg•  NASA Instrument Cost Model (NICM) for primary payload

instruments3.  Applies existing best practices to adjust for system level cost

considerations (e.g. learning curve, design heritage)4.  Uses current launch vehicle market prices to estimate launch

cost and operational support requirements 5.  Formats cost estimate and record caveats to valuation

•  Shao et al. (2014) took a similar approach to Performance-Based Cost Modeling, leveraging USCM, SSCM, NICM

Future work: Weighting function at crossover

11

Sample Output

{ "constellationCost": { "totalCost": { "estimate": 285896.029, "standardError": null, "confidenceInterval": [lowLimit, highLimit, probability] "caveats": "Constellation is homogeneous. Launch Vehicle was not designated, launch vehicle cost is set to 0. " }, "rdteCost": { "estimate": 81346.16106, "standardError": null, "confidenceInterval": [lowLimit, highLimit, probability] "caveats": "CER choice: Input (spacecraft total mass) to thermal RDT&E CER 2 for spacecraft 1 is out of acceptable CER range. CER 1 was used instead. " }, "drivers": "Spacecraft 1, Payload. Spacecraft 1, Operations. Spacecraft 2 IA&T. ", "spacecraftRank": [1,2] }}

Truncated output JSONAdvantages: -  Human readable, promotes transparency-  Formatted to mimic Work Breakdown

Structure-  Interoperability

12

Sample Output


13

Sample Output


14

Sample Output


15

Sample Output


16

Sample Output

{ "constellationCost": { "totalCost": { "estimate": 285896.029, "standardError": null, "confidenceInterval": [lowLimit, highLimit, probability] "caveats": "Constellation is homogeneous. Launch Vehicle was not designated, launch vehicle cost is set to 0. " }, "rdteCost": { "estimate": 81346.16106, "standardError": null, "confidenceInterval": [lowLimit, highLimit, probability] "caveats": "CER choice: Input (spacecraft total mass) to thermal RDT&E CER 2 for spacecraft 1 is out of acceptable CER range. CER 1 was used instead. " }, "drivers": "Spacecraft 1, Payload. Spacecraft 1, Operations. Spacecraft 2 IA&T", "spacecraftRank": [1,2] }}

17

Sample Output

{ "constellationCost": { "totalCost": { "estimate": 285896.029, "standardError": null, "confidenceInterval": [lowLimit, highLimit, probability] "caveats": "Constellation is homogeneous. Launch Vehicle was not designated, launch vehicle cost is set to 0. " }, "rdteCost": { "estimate": 81346.16106, "standardError": null, "confidenceInterval": [lowLimit, highLimit, probability] "caveats": "CER choice: Input (spacecraft total mass) to thermal RDT&E CER 2 for spacecraft 1 is out of acceptable CER range. CER 1 was used instead. " }, "drivers": "Spacecraft 1, Payload. Spacecraft 1, Operations. Spacecraft 2 IA&T", "spacecraftRank": [1,2] }}

18

Current Status

•  Cost Routine Version 1 is being integrated TAT-C as a set of MATLAB functions

•  Short term remaining tasks:–  Transition model to C++–  Cost Risk Estimation, will depend on risk methodology–  Operations and Ground Segment

•  Future Work

•  Long term: –  Continued bench marking the model for reliability–  Upgrade CERs to most recent formulae

19

Future Work

•  Operational Challenges–  Operations and maintenance can be most expensive

constellation mission element–  How reliable are existing methods for constellations

and what is the impact of increasing automation?• Automation would increase Research and Development

but decrease cost of operations

•  Discontinuity at transition between USCM and SSCM –  Weighting function (0-1) applied near transition–  Will determine the appropriate blend of these methods

during final model benchmarking

20

Conclusion

•  Aggregate cost model leverages existing tools and applies them to DSM and constellation architectures, while addressing limitations of existing methodologies

•  This approach allows for integration of cost with early tradespace exploration–  VCR is designed for TAT-C, but the form and function will

allow for interoperability –  Promote cost estimating transparency in automated

processes–  Relative cost estimates for architecture comparison

•  Continues to reveal limits of cost estimating techniques for constellations, and inform future DSM development

21

References and Acknowledgements

1.  Wertz, James Richard., and Wiley J. Larson. Space Mission Analysis and Design. 3rd ed. Torrance, CA: Microcosm, 1999. Print.2.  Bearden, David A. "Small-satellite costs." Crosslink 2.1 (2001): 32-44.3.  “TROPICS: Mission Overview." Mission Overview. MIT Lincoln Laboratory, 2015. Web. 1 June 2016.4.  Tieu, B., J. Kropp, and N. Lozzi. "The Unmanned Space Vehicle Cost Model-Past, Present, and Future." American Institute of

Aeronautics and Astronautics: Space 2000 Conference and Exposition. Long Beach, California. 2000.5.  U.S. Geological Survey, 2015, Landsat—Earth observation satellites: U.S. Geological Survey Fact Sheet 2015–3081, 4 p.,

http://dx.doi.org/10.3133/fs20153081.6.  Nguyen, Phu, et al. "Unmanned Space Vehicle Cost Model." US Air Force Space and Missile Systems Center (SMC/FMC) 2430

(1994): 90245-4687.7.  Lao, N. Y., T. J. Mosher, and J. M. Neff. "Small Satellite Cost Model, Version 98 InTRO." (1999).8.  Swartwout, M. "Cubesats by Mission Type." Cubesat Database. N.p., 08 June 2016. Web. 9.  Hamaker, Joseph, and Ronald Larson. "Quick Cost 6.0: Introduction and Overview." NASA 2015 Cost Symposium. Ames

Research Center, Moffett Field, CA. NASA 2015 Cost Symposium Presentations. NASA. Web. 5 Oct. 2015.10. Burgess, Eric, Darren Elliot, and Charles Hunt. "Programmatic Estimating Tool: Parametric-based Cost, Schedule, & Phasing

Health Check." NASA 2015 Cost Symposium. Ames Research Center, Moffett Field, CA. NASA 2015 Cost Symposium Presentations. Web. 5 Oct. 2015.

11. Habib-Agahi, Hamid, Joe Mrozinski, and George Fox. "NASA instrument cost/schedule model Hamid Habib-Agahi." Aerospace Conference, 2011 IEEE. IEEE, 2011.

12. NASA, NASA. "Cost Estimating Handbook." Washington, DC (2008).13. Abramson, R. L., and D. A. Bearden. "Cost Analysis Methodology for High-Performance Small Satellites." SPIE International

Symposium on Aerospace and Remote Sensing, Small Satellite Technology and Applications III. 1993.14. Nag, Sreeja, Jacqueline LeMoigne, and Olivier de Weck. "Cost and risk analysis of small satellite constellations for earth

observation." Aerospace Conference, 2014 IEEE. IEEE, 2014.15. Shao, Anthony, Elizabeth A. Koltz, and James R. Wertz. "Performance Based Cost Modeling: Quantifying the Cost Reduction

Potential of Small Observation Satellties." AIAA Reinventing Space Conference, AIAA-RS-2013-1003, Los Angeles, CA, Oct. 2013.Bearden, David A. “Small Satellite Costs.” Crosslink. Aerospace Corporation, 2000. Web. 6 Jun. 2016.

This material is based on work supported by the National Science Foundation Graduate Research Fellowship under Grant No. 1122374. Any opinion, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.

Guidance and support have also been provided by the Massachusetts Institute of Technology and NASA Goddard Space Flight Center.

22

Thank you for your attention! Any questions?

23

Additional Slides: Key Assumptions

•  Comparative, not an exact value, estimate–  Estimate should provide an approximation that can be used

for tradespace analysis •  Comparison during concept evaluation, not as direct

budgeting tool •  CERs are based in historical trends; assume that the

trends will hold into the foreseeable future –  Major technological changes will impact model fidelity–  Smallsat launchers could cause significant changes

•  Prototype, not protoflight, hardware development process •  Scope creep is not considered •  Project is executed at the optimal pace

24

Acronyms

•  CER: Cost Estimating Relationship•  COTS: Commercial Off The Shelf•  DSM: Distributed Spacecraft Missions•  DOD: Department of Defense•  IR: Infrared •  JSON: JavaScript Object Notation•  NASA: National Aeronautics and Space Administration•  NICM: NASA Instrument Cost Model•  PET: Programmatic Estimating Tool•  RDT&E: Research Development, Test, & Evaluation •  SSCM: Small Satellite Cost Model •  TAT-C: Tradespace Analysis Tool for Constellations•  USCM: Unmanned Space Vehicle Cost Model •  VCR: Value, Cost, and Risk

satellite constellation cost modeling: an aggregate model · 6/14/2016 · 1 satellite...

Documents